Update app.py

app.py

@@ -1,47 +1,59 @@
 import io
 import cv2
 import numpy as np
+import requests
+import os
 from fastapi import FastAPI, UploadFile, File
 from fastapi.responses import Response
-from PIL import Image
 from realesrgan import RealESRGANer
 from basicsr.archs.rrdbnet_arch import RRDBNet
 
 app = FastAPI()
 
-# 🧠 Load the Real-ESRGAN Model (The powerful AI)
+# 🧠 Setup Real-ESRGAN Model
+model_url = 'https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth'
+model_path = 'RealESRGAN_x4plus.pth'
+
+if not os.path.exists(model_path):
+    print("Downloading AI model...")
+    response = requests.get(model_url)
+    with open(model_path, 'wb') as f:
+        f.write(response.content)
+
 model = RRDBNet(num_in_ch=3, num_out_ch=3, num_feat=64, num_block=23, num_grow_ch=32, scale=4)
 upsampler = RealESRGANer(
     scale=4,
-    model_path='https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth',
+    model_path=model_path,
     model=model,
-    tile=400, # Tile processing to prevent memory crash
+    tile=400, # Prevents "Out of Memory" errors on free CPUs
     tile_pad=10,
     pre_pad=0,
-    half=False # Keep False for HF CPU spaces
+    half=False
 )
 
 @app.get("/")
 def home():
-    return {"status": "AI Super-Resolution Online"}
+    return {"status": "Silent Neural HD Engine Online"}
 
 @app.post("/upscale")
 async def upscale(file: UploadFile = File(...)):
-    # 1. Read uploaded file
-    data = await file.read()
-    nparr = np.frombuffer(data, np.uint8)
-    img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
-    
-    # 2. AI Neural Inference
     try:
-        # This is where the actual AI magic happens
+        # Read image
+        data = await file.read()
+        nparr = np.frombuffer(data, np.uint8)
+        img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
+        
+        # Actual AI Processing
+        # outscale=4 makes it 4x bigger and sharper
         output, _ = upsampler.enhance(img, outscale=4)
         
-        # 3. Encode result
-        _, encoded_img = cv2.imencode('.jpg', output, [int(cv2.IMWRITE_JPEG_QUALITY), 95])
+        # Convert back to JPG
+        _, encoded_img = cv2.imencode('.jpg', output, [int(cv2.IMWRITE_JPEG_QUALITY), 90])
         return Response(content=encoded_img.tobytes(), media_type="image/jpeg")
+    
     except Exception as e:
-        return {"error": str(e)}
+        print(f"Error: {str(e)}")
+        return {"error": "AI Engine busy or image too large"}
 
 if __name__ == "__main__":
     import uvicorn